Lovable hits $300M ARR: co-founder Anton Osika on building a technical co-founder for the masses

Summary

Lovable, an AI-powered app creation platform for non-technical users, reached $300M ARR just two months after hitting $200M. Co-founder and CEO Anton Osika attributes the acceleration to users actually building businesses with the platform rather than treating it as a novelty.

A restaurant software provider built a voice AI phone system over a weekend that handles customer calls. Someone else is shipping AI-generated dog pictures in frames and generating $300K monthly revenue. A healthcare AI staffing company bootstrapped to $1M ARR in five months. These use cases show Lovable functioning as a technical co-founder for people who would otherwise never build software.

Enterprise adoption is now scaling. One real estate software company with 80,000+ agents across dozens of countries needed to replatform hundreds of websites on a one-year timeline. Using Lovable, they finished in three weeks and saved $2M by eliminating a separate AI chatbot vendor. The company sees 300M visits to Lovable-built applications per month, mostly organic word-of-mouth as users discover apps and ask how they were built.

The maintenance problem is becoming real. It's now trivial to spin up 200 internal tools, but expensive to maintain them. Osika argues the problem isn't the platform but the approach. Users should move fast to validate what works, then loop back with the AI to rethink whether incremental maintenance makes sense or the system should be rebuilt from first principles. A recent product update added advanced planning mode to prompt users about their actual objectives before building, reducing maintenance-heavy implementations.

Poorly specified prompts create what Osika calls "vibe coding debt." Rather than constrain speed upfront, his philosophy is to build fast and then invest in deliberate refactoring, treating the AI as an ongoing design partner.

Security and quality gates matter at scale. As Lovable-built apps reach millions of users, the platform has spent the last nine months hardening security. Someone needs to own quality before apps reach production, even if multiple team members can tinker and submit changes. This is tacit acknowledgment that frictionless creation creates risk when deployed by non-engineers at enterprise scale.

Osika doesn't track competitors or read other companies' job postings. Instead, Lovable focuses on first principles: what is the best interface for humans to interact with AI systems as capability grows? That orientation sometimes means not adding features, favoring simplicity and capability density over sprawl. Users consistently choose Lovable for simplicity.

The team is hiring aggressively, including at Anthropic. If you use AI correctly as a business tool, each person outputs more. Lovable is looking for curious people who "get things and understand them quickly," not necessarily software engineers.

Net dollar retention exceeds 100%, though some users experiment once and churn. The company is segmenting by jobs to be done to understand where to improve core experience. Retention improves once users build something tangible like a website or internal tool. The comparison is to website builders: hosting and ongoing maintenance lock users in.

Partnerships over vertical integration. Lovable recommends Supabase, Stripe, and other infrastructure providers when users need them. Rather than build integrations in-house, Osika wants to deepen partnerships. Some partners are seeing majority growth from Lovable's recommendations, and as ease-of-use improves, Lovable will naturally encroach on those vendors.

Growth is driven by word-of-mouth and social proof. Twitter and LinkedIn are primary channels, with Instagram on the radar as a mainstream lever. The 300M monthly visits to Lovable apps is the real engine: users discover a well-designed site or tool, learn it was built with Lovable, and want to build themselves.

Lovable uses a combination of models from different labs rather than betting on one. This lets the team optimize for user experience without being hostage to any single provider's roadmap. The team has been experimenting with voice-based agents inspired by Claude Bot and personal assistants that can talk to all internal tools and pull data on demand. Agent-based team collaboration is a top goal for the year.

As AI capability grows, competitive advantage shifts from access to models to simplicity, security, and trustworthiness of the interface. Lovable started with non-technical users and designed down to simplicity, which turns out to work better even for sophisticated users than tools built by developers for developers.